Photovoltaic (PV) cells are used to convert solar energy (sunlight) into electricity, and are typically implemented either in flat-panel arrangement, or in conjunction with concentrating solar collectors.
Solar Energy Harvesting Requires Inexpensive Large Area Components
Solar energy arrives at the surface of the earth as a relatively dilute form of radiant energy, peaking at approximately 1000 W/m2. Any solar energy harvesting system is therefore required to cover a relatively large area in order to intercept enough sunlight for a meaningful power output. The intercepting area can consist of the energy converting components themselves (e.g., photovoltaic cells in a flat panel module) or consist of optical elements used to direct the intercepted light to a typically smaller converting component (e.g., a higher performance photovoltaic cell in a solar concentrator system). Due to the low price of electricity to which industrialized nations have become accustomed, the key techno-economic challenge and driver is to make the solar energy harvesting system very inexpensive per unit area.
Fabrication processes for photovoltaic cells (PV cells) have benefited from the mature status and sustained progress in semiconductor manufacturing techniques developed for the field of microelectronics. Although it can be expected that process improvements will continue to lower the cost of PV cells into the future, the often cited analogy with Moore's law in microelectronics is only partially appropriate: Moore's law rests heavily on a reduction in surface area per useful unit (e.g. a transistor), while the useful unit in a PV cell is surface area itself. The usefulness of the surface area can be modified in a first example by improving the efficiency of the PV cell, which—being an efficiency metric—naturally has fundamental limits forcing the progress trajectory into an S-curve, and is not the objective of this invention. The usefulness of the surface area can be modified in a second example by optical concentration. It is the objective of this invention to achieve moderate concentration levels (e.g., 10× to 40×, or sometimes higher) with system components that can scale to very low cost and do not incur the system disadvantages typically associated with conventional solar concentrators.
Benefits of Solar Concentration
Historically, a single one of the beneficial aspects of low/medium concentration PV systems dominated the discussion: this dominant aspect was the paradigm of “saving silicon”, which cannot be the only motivation anymore in times of low cost silicon feedstocks now available for PV cell production. Whether flat panel PV cells or PV cells working in somewhat concentrated light environments will form the mainstay of our futures solar energy systems is still heavily debated today. Extrapolations are subject high uncertainty due to the industrial network effects unfolding over time. While flat panel approaches may well win the race to grid parity, there are some important arguments to note in favor of concentrated approaches. These benefits are given here with a bias towards low/medium concentrators and receivers of similar complexity to crystalline silicon PV cells:
(1) A solar harvesting device having only a portion of the surface area consisting of fragile, electrically connected PV cells allows more degrees of freedom in the system design. These can be used to make the system more robust, flexible, easier to ship, partially translucent, building integrated, just to name a few possibly directions. An economic degree of freedom is won by the fact, that a low/medium concentrator can make good use of higher efficiency cells (e.g. high efficiency silicon or similar), that might be temporarily or systematically not quite competitive for flat panel application under lx sun.
(2) A large fraction of the capital expenditure of PV module manufacturing goes towards the fabrication machinery for PV cell manufacturing. Producing more total PV module area per year normally requires the installation of proportionally more machinery. This can limit the growth of such a technology and business domain, as the capital for expansion often is the limiting resource. Organic growth from reinvesting profits may be too slow for staying on top of competitors in the market or climate goals in the world. A concentration system can mitigate such capital imposed limits and allow faster scale-up in terms of GWp/year, since the electrical productivity of each cell leaving the (capacity limited) production line is increased by the flux concentration factor F=Cg*η (where η is the optical efficiency of the concentrator). Of course, this strategy only holds if the capital cost to obtain fabrication capacity for the optical concentrator is lower on a per Watt basis than the PV cell fabrication facility (fab). This is particularly true for optical concentration elements that can be manufactured on existing machinery. As explained below, systems described in this invention can in large parts be manufactured on film or foil manufacturing equipment, most readily on the microoptical film fabrication equipment that exists today for display backlighting films. These plants have an annual capacity in the hundreds of square kilometers per year. The ability to rapidly scale up production will become a particularly prominent competitive differentiator, as soon as solar energy first undercuts the costs of established fossil fuel based generation.
(3) Concentrator photovoltaics can drastically reduce the embodied energy in the solar energy harvesting system per peak Watt installed when compared to flat panel PV cells. This shortens the energy payback time of the system and similarly the “energy returned on energy invested”.
(4) Returning to the initial point made above, a low/medium solar concentrator with good manufacturability can reduce the cost per installed W compared to flat panel systems, if disadvantages of prior art concentrators (such as tracking requirements) are avoided.
Passive Optical Concentrators in the Prior Art
Prior art solar concentrators utilize optics (e.g., reflectors, lenses, etc.) to focus sunlight onto a relatively small PV cell. This can be motivated by direct cost savings (e.g., when the area specific cost of the optics is lower than the cost of the PV cell), and/or by the desire for higher system efficiencies (e.g., by allowing to use high performance PV cells that are only available and economic in small areas).
Prior art approaches have so far been mostly passive optical systems, which are defined here as systems that do not substantially change the wavelength of the light they process. In passive optical systems, concentration in the spatial domain comes at the expense of an expansion in the angular domain. This is mandated by principles of conservation of phase space (i.e., Etendue).
The concentration sought from a solar concentrator is a concentration in the spatial domain: The energy intercepted at a large area aperture is coupled to a small area receiver (photovoltaic or thermal) having a surface area that is smaller by a factor Cg. This causes the solid angular subtended by the incoming radiation to expand by approximately the same factor (modified by the refractive index contrast and projection direction) before it reaches the smaller receiver. However, the solid angle from which a receiver can accept light is typically limited to 2π2 (hemispherical space) or in some cases to the absolute limit of the full sphere at 4π2. This limits the solid angle from which a concentrator can efficiently accept incoming radiation at its input. However, even direct sunlight originates over the course of year from within a significant portion of the sky hemisphere. The acceptance solid angle starts to become restricted to a solid angle zone narrower than this even for very low spatial concentration factors Cg, e.g. 3×. This can be improved upon by optimizing for the particular angular intensity distribution, but passive static systems beyond 10× concentration are impractical on earth.
It should be noted that the direct sunlight itself subtends only a very small solid angle at any given time. Based on this, prior art systems are able to efficiently reach higher concentration factors by going from static (untracked) systems to tracked concentrators. These tracking systems keep the relative angular position between the sun and the concentrator substantially constant in one or two of the angular dimensions; typically by mechanical movement of the systems. Mechanical tracking systems add installation cost, maintenance cost, reliability concerns, windloading problems and other disadvantages to the system. A system that achieves higher concentration factors than static concentrators without mechanical tracking is therefore highly desirable.
Luminescent Solar Concentrators in the Prior Art
A Luminescent Solar Concentrator (LSC) allows concentration without tracking of both diffuse and direct radiation and have been described in the prior art. LSCs overcome the single wavelength Etendue limits that constrain passive optical concentrators by subjecting each photon to a downward shift in energy (towards longer wavelength), e.g. via a fluorescence process. The photon energy difference is required for compliance with the governing thermodynamic principles and enables concentration factors well beyond the domain to which static concentrators are limited otherwise.
FIG. 13 is a simplified side view illustrating a prior art luminescent concentrator 40, which represents another type of concentrating solar collector (see, e.g., U.S. Pub. App. 2009/0235974, which is incorporated herein by reference in its entirety). Luminescent concentrator 40 includes a light-guiding slab 41 containing a dye layer 46, and one or more PV cells 50 that are secured to light-guiding slab 41. Light-guiding slab 41 is formed using a light transparent material (e.g., glass or plastic) and includes opposing upper and lower planar “broadside” surfaces 42 and 44, and side edges 45 that extend between upper surface 42 and lower surface 44. Dye layer 46 is disposed between upper surface 42 and lower surface 44, and includes one or more luminescent (e.g., fluorescent) materials capable of absorbing sunlight SL (indicated by parallel dashed line arrows in FIG. 13) having a first wavelength, and generating light emissions LE (indicated by dash-dot line arrows in FIG. 13) that have a second, typically longer wavelength. In an alternative embodiment, a separate dye layer is omitted, and the luminescent material is uniformly dispersed throughout slab 41 such that it extends between upper surface 42 and lower surface. PV cells 50 are sized to fit along side edges 45 of light-guiding slab 41. As illustrated by the paths of light emissions LE in FIG. 13, the light transparent material of light-guiding slab 41 is selected such that, during operation, upper surface 42 and lower surface 44 cause a large fraction of the light emissions from the luminescent materials to be retained within light-guiding slab 41 (i.e., light emissions LE are guided between upper surface 42 and lower surface 44 by way of total internal reflection). The fraction of thereby retained light in this step immediately after emission of the fluorophore is often referred to as “trapping efficiency” and is depending on the refractive index contrast between the light guiding slab and its immediate surroundings; with typical values being around 75% (for this step alone) in the case of a slab with refractive index n=1.5 in air, for the case of approximately isotropic emission. The retained light emissions LE travel through light-guiding slab 41 until they encounter typically a side edge 45, at which point the light emissions exit light-guiding slab 41 and are absorbed by PV cells 50, which convert the light emissions into electrical energy. Luminescent concentrator 40 provides an advantage over optical concentrators in that it provides very higher optical concentration factors than otherwise possible without mechanical tracking.
Core Problems of Prior Art LSCs
While LSC hold great promise for future generations of PV systems, two major challenges have blocked their path to success so far. The first is a materials issue (not addressed with this invention, but solved elsewhere). The second problem is an unfavorable scaling behavior of the optical efficiency and/or thickness with area of the collecting and lightguiding slab (addressed with this invention). The former materials related issue concerned the lifetime of suitable luminescent materials. Historically, organic dyes were employed exhibiting insufficient lifetimes under sunlight conditions. Recently however, sunlight stable dyes (such as BASF Lumogen 305 dyes) have become available and are commercially marketed for solar energy applications with good lifetimes in experiments and product specifications. At the same time, quantum dots and newer phosphors have been found and applied to luminescent solar concentrators. Their inorganic composition circumvents the lifetime concerns of dyes altogether.
Transport Losses as the Main Scaling Impediment in Prior Art LSCs
The latter scaling issue arises from the quantitative coupling between lightguide thickness t, chemical concentration of the luminescent material in the slab c, receiver spacing s, and receiver size r. Size r and spacing s are proportional to each other for a fixedly chosen optical concentration factor Cg.
The average pathlength which a photon needs to be transported inside the lightguide slabs scales with spacing s. Transport losses are mainly due to reabsorption in the luminescent material and increase as the transported distance increases with spacing s. This would motivate to reduce spacing (and consequently size r). However, very small PV cells used as receivers (small size r, at least in one dimension) are economically strongly discouraged. They are expensive to make (due to dicing and processing cost), less efficient (due to edge passivation and/or introduced shunt resistances), costly in assembly (due to handling and higher number of required electrical connections) and tend to have yield and reliability issues (due to more situations of possible breakage).
In prior art LSCs, overly small receivers can only be avoided by accepting longer transport distances. To then avoid excessive transport losses, the concentration of luminescent material c (determining the absorbance for both initial sunlight capture and transport reabsorption) must be lowered in an inversely proportional fashion to compensate. However, this in turn necessitates an increased thickness of the lightguide slab, since impinging sunlight would otherwise excessively transmit through the slab uncaptured by the luminescent material. A thicker lightguide slab (being the large area component in the system) would increase system volume, weight and ultimately mass manufactured cost accordingly and is therefore not a fruitful scaling route to cover order of magnitude changes.
It should be noted that the concentration c above is given for a luminescent material which is homogenously dispersed throughout the lightguiding slab. However, the same argument holds to first order with the same scaling behavior for luminescent material placed as one ore more layers, or inhomogeneously distributed on or in the lightguiding slab. In such cases, c refers to the average concentration along the thickness direction of the slab. This holds because of the multiplicative interchangability of thickness and concentration in the exponent of the Lambert Beer law, which is used as a good approximation to the typical transmission decay observed (e.g., by Bode) in LSC materials.
What is needed is a low-cost solar concentrator that provides the advantages of a luminescent concentrator, but avoids the problems associated with existing prior art luminescent concentrators. What is also needed is a solar energy harvesting system utilizing such a low-cost, luminescent concentrator.